Aerobic oxidation of primary alcohols (including methanol) by copper(II)- and zinc(II)-phenoxyl radical catalysts

Article abstract of DOI:10.1021/ja991481t

The tetradentate ligand N,N′-bis(3,5-di-tert-butyl-2-hydroxyphenyl)-1,2-phenylenediamme, H₄L¹, has been prepared, and its square planar complexes [Cu¹¹(L³)] and [Zn¹¹(L³)] have been synthesized from the reaction of H₄L¹ with [Cu¹(NCCH₃)₄](CIO₄) or Zn(BF₄)₂·2H₂O in methanol in the presence of air. The dianion (L³)^2- represents the two-electron oxidized form of (L¹)^4-, namely N,N′-bis(3,5-di-ferf-butyl-2-hydroxyphenyl)-1,2-diiminoquinone. Complexes [Cu¹¹(L³)]·CH₃CN and [Zn(L³)]·CH₃CN have been characterized by X-ray crystallography, EPR spectroscopy, and magnetochemistry; [Cu¹¹(L³)] has an S = 1/2 ground state, and [Zn-(L³)] is diamagnetic. Cyclic voltammetry established that both complexes undergo two successive reversible one-electron oxidations and two successive reversible one-electron reductions. Thus, the coordinated ligand exists in five oxidation levels. The species [M¹¹(L⁴)]PF₆ (M = Cu¹¹, Zn¹¹) and [M¹¹(L⁴)](ClO₄)₂ (M = Cu¹¹, Zn¹¹) have been isolated and characterized by UV/vis, EPR, and ¹H NMR spectroscopy and by magnetic susceptibility measurements, where (L⁴)^- represents the monoanion N-(3,5-di-tert-butyl-2-hydroxyphenyl)-N′-(3,5-di-tert-butyl-2-phenoxyl)-1, 2-diiminoquinone and (L⁵) is the neutral ligand N,N′-bis(3,5-di-tert-butyl-2-phenoxyl)-1,2-diiminoquinone. Similarly, two complexes of the type [M¹¹(L¹H₂)] (M = Cu¹¹, Zn¹¹) have been isolated from the reaction of L¹H₄ with Cu¹¹(ClO₄)₂·6H₂O or Zn(ClO₄)₂·6H₂O under anaerobic conditions in the presence of NEt₃. Complexes [Cu¹¹(L₄)]PF₆ and [Zn(L₄)]PF₆ selectively oxidize primary alcohols (including methanol and ethanol) in a stoichiometric fashion under anaerobic conditions, yielding the corresponding aldehydes and [M¹¹(L²H₂)]⁺ (M = Cu¹¹, Zn¹¹), where (L²)^3- is the trianionic form of N,N′-bis(3,5-di-fert-butyl-2-hydroxyphenyl)-l,2-diiminosemiquinone. Since the latter reduced forms react rapidly with dioxygen with formation of [M¹¹(L⁴)]⁺ (M = Cu, Zn) and 1 equiv of H₂O₂, these oxidized species are catalysts for the air oxidation of primary alcohols, including ethanol and methanol, with concomitant formation of H₂O₂ and aldehydes. The kinetics of the stoichiometric reactions and of the catalyses (initial rate method) have been measured. Large kinetic isotope effects show that H-abstraction from the α-carbon atom of a coordinated alcoholato ligand is the rate-determining step in all cases.

Full text of DOI:10.1021/ja991481t

J. Am. Chem. Soc. 1999, 121, 9599-9610
9599
Aerobic Oxidation of Primary Alcohols (Including Methanol) by
Copper(II)- and Zinc(II)-Phenoxyl Radical Catalysts
Phalguni Chaudhuri,* Martina Hess, Jochen Mu1ller, Knut Hildenbrand, Eckhard Bill,
Thomas Weyhermu1ller, and Karl Wieghardt*
Contribution from the Max-Planck-Institut fu¨r Strahlenchemie, Stiftstrasse 34-36,
D-45470 Mu¨lheim a.d. Ruhr, Germany
ReceiVed May 5, 1999
Abstract: The tetradentate ligand N,N′-bis(3,5-di-tert-butyl-2-hydroxyphenyl)-1,2-phenylenediamine, H4L1, has
been prepared, and its square planar complexes [CuII(L3)] and [ZnII(L3)] have been synthesized from the reaction
of H4L1 with [CuI(NCCH3)4](ClO4) or Zn(BF4)2‚2H2O in methanol in the presence of air. The dianion (L3)2-
represents the two-electron oxidized form of (L1)4-, namely N,N′-bis(3,5-di-tert-butyl-2-hydroxyphenyl)-1,2-
diiminoquinone. Complexes [CuII(L3)]‚CH3CN and [Zn(L3)]‚CH3CN have been characterized by X-ray
1
crystallography, EPR spectroscopy, and magnetochemistry; [CuII(L3)] has an S ) /2 ground state, and [Zn-
(L3)] is diamagnetic. Cyclic voltammetry established that both complexes undergo two successive reversible
one-electron oxidations and two successive reversible one-electron reductions. Thus, the coordinated ligand
exists in five oxidation levels. The species [MII(L4)]PF6 (M ) CuII, ZnII) and [MII(L5)](ClO4)2 (M ) CuII,
1
ZnII) have been isolated and characterized by UV/vis, EPR, and H NMR spectroscopy and by magnetic
susceptibility measurements, where (L4)- represents the monoanion N-(3,5-di-tert-butyl-2-hydroxyphenyl)-
N′-(3,5-di-tert-butyl-2-phenoxyl)-1,2-diiminoquinone and (L5) is the neutral ligand N,N′-bis(3,5-di-tert-butyl-
2-phenoxyl)-1,2-diiminoquinone. Similarly, two complexes of the type [MII(L1H2)] (M ) CuII, ZnII) have
been isolated from the reaction of L1H4 with CuII(ClO4)2‚6H2O or Zn(ClO4)2‚6H2O under anaerobic conditions
in the presence of NEt3. Complexes [CuII(L4)]PF6 and [Zn(L4)]PF6 selectively oxidize primary alcohols (including
methanol and ethanol) in a stoichiometric fashion under anaerobic conditions, yielding the corresponding
aldehydes and [MII(L2H2)]+ (M ) CuII, ZnII), where (L2)3- is the trianionic form of N,N′-bis(3,5-di-tert-butyl-
2-hydroxyphenyl)-1,2-diiminosemiquinone. Since the latter reduced forms react rapidly with dioxygen with
formation of [MII(L4)]+ (M ) Cu, Zn) and 1 equiv of H2O2, these oxidized species are catalysts for the air
oxidation of primary alcohols, including ethanol and methanol, with concomitant formation of H2O2 and
aldehydes. The kinetics of the stoichiometric reactions and of the catalyses (initial rate method) have been
measured. Large kinetic isotope effects show that H-abstraction from the R-carbon atom of a coordinated
alcoholato ligand is the rate-determining step in all cases.
Introduction
reaction from the R-carbon atom of the alcoholate to the
coordinated tyrosyl radical, generating a bound ketyl radical
anion and a tyrosine. The ketyl-radical ligand is then intramo-
lecularly converted via a one-electron oxidation to the aldehyde
with concomitant formation of a CuI species. Reoxidation of
the latter species by O2 regenerates the active CuII-tyrosyl form
and H2O2 (Scheme 1).
The copper-containing metalloprotein galactose oxidase (GO)
is an enzyme which catalyzes selectively the aerobic oxidation
of primary alcohols to aldehydes with formation of 1 equiv of
hydrogen peroxide (eq 1). In the active site of the active form
RCH2OH + O2 [GO]8 RCHO + H2O2
(1)
(3) (a) Wachter, R. M.; Montagne-Smith, M. P.; Branchaud, B. P. J.
Am. Chem. Soc. 1997, 119, 7743. (b) Branchaud, B. P.; Montagne-Smith,
M. P.; Kosman, D. J.; McLaren, F. R. J. Am. Chem. Soc. 1993, 115, 798.
(c) Whittaker, M. M.; Whittaker, J. M. Biophys. J. 1993, 64, 762. (d)
Whittaker, M. M.; Ballou, D. P.; Whittaker, J. W. Biochemistry 1998, 37,
8426. (e) Wachter, R. M.; Branchaud, B. P. J. Am. Chem. Soc. 1996, 118,
2782. (f) Whittaker, J. W.; Whittaker, M. M. Pure Appl. Chem. 1998, 70,
903.
(4) (a) Halfen, J. A.; Jazdzewski, B. A.; Mahapatra, S.; Berreau, L. M.;
Wilkinson, E. C.; Que, L., Jr.; Tolman, W. B. J. Am. Chem. Soc. 1997,
119, 8217. (b) Sokolowski, A.; Leutbecher, H.; Weyhermu¨ller, T.; Schnepf,
R.; Bothe, E.; Bill, E.; Hildebrandt, P.; Wieghardt, K. J. Biol. Inorg. Chem.
1997, 2, 444. (c) Zurita, D.; Gautier-Luneau, I.; Menage, S.; Pierre, J.-L.;
Saint-Aman, E. J. Biol. Inorg. Chem. 1997, 2, 46. (d) Itoh, S.; Takayama,
S.; Arakawa, R.; Furuta, A.; Komatsu, M.; Ishida, A.; Takamuku, S.;
Fukuzumi, S. Inorg. Chem. 1997, 36, 1407. (e) Zurita, D.; Menage, S.;
Pierre, J.-L.; Saint-Aman, E. New J. Chem. 1997, 21, 1001. (f) Mu¨ller, J.;
Weyhermu¨ller, T.; Bill, E.; Hildebrandt, O.; Ould-Moussa, L.; Glaser, T.;
Wieghardt, K. Angew. Chem., Int. Ed. 1998, 37, 616. (g) Halcrow, M. A.;
Chia, L. M. L.; Liu, X.; McInnes, J. L.; Yellowlees, L. J.; Mabbs, F. E.;
Davies, J. E. Chem. Commun. 1998, 2465.
of this enzyme, a single copper(II) ion is coordinatively bound
to a (thioether-modified) tyrosyl radical,1 two histidine residues,
and a tyrosinate anion (or tyrosine).2
The mechanism of the catalytic conversion of primary
alcohols by GO is well understood.3 In the rate-determining step,
an O-coordinated alcoholate ligand undergoes an H-abstraction
(1) (a) Whittaker, J. W. In Metal Ions in Biological Systems; Sigel, H.,
Sigel, A., Eds.; Marcel Dekker: New York, 1994; Vol. 30, pp 315-360.
(b) Knowles, P. F.; Ito, N. In PerspectiVes in Bio-inorganic Chemistry; Jai
Press Ltd.: London, 1994; Vol. 2, pp 207-244.
(2) (a) Ito, N.; Phillips, S. E. V.; Stevens, C.; Ogel, Z. B.; McPherson,
M. J.; Keen, J. N.; Yadav, K. D. S.; Knowles, P. F. Nature 1991, 350, 87.
(b) Ito, N.; Phillips, S. E. V.; Stevens, C.; Ogel, Z. B.; McPherson, M. J.;
Keen, J. N.; Yadav, K. D. S.; Knowles, P. F. Faraday Discuss. 1992, 93,
75. (c) Ito, N.; Phillips, S. E. V.; Yadav, K. D. S.; Knowles, P. F. J. Mol.
Biol. 1994, 238, 794.
10.1021/ja991481t CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/04/1999

9600 J. Am. Chem. Soc., Vol. 121, No. 41, 1999
Chaudhuri et al.
R2CHOH + O2 [B]8 R2CdO + H2O2
(2)
Scheme 1. Proposed Reaction Mechanism of GO, Ref 3
2R2CHOH + O2 [B]8 R2C(OH)-C(OH)R2 + H2O2 (3)
Finally, the mononuclear complex C7 oxidizes only primary
alcohols, including ethanol, by dioxygen to aldehydes and H2O2,
and D has been reported to catalyze the electrochemical
oxidation of alcohols.5c
Here we report two new catalysts which effectively oxidize
primary alcohols, including methanol, with dioxygen to alde-
hydes and H2O2 but not secondary alcohols, such as 2-propanol
or diphenylcarbinol. Cu complexes capable of catalyzing the
oxidation of alcohols to aldehydes with air have been described
previously, but water and not H2O2 is the reduction product.8
We have synthesized the new tetradentate ligand N,N′-bis-
(3,5-di-tert-butyl-2-hydroxyphenyl)-1,2-phenylenediamine, H4-
(L1), and investigated its coordination chemistry with Cu and
Zn ions. This ligand exists in five oxidation levels, as shown in
Scheme 2; Chart 2 lists the complexes prepared, with their labels
and spin ground states.
Chart 1. Catalysts for the Aerobic Oxidation of Alcohols
Results
Syntheses and Characterization of Complexes. The ligand
N,N′-bis(3,5-di-tert-butyl-2-hydroxyphenyl)-1,2-phenylenedi-
amine, H4(L1), has been prepared by a condensation reaction
of 3,5-di-tert-butylcatechol and o-phenylenediamine (2:1) in the
presence of triethylamine in n-heptane. Scheme 2 shows one
resonance structure for each of the five oxidation states of this
ligand (L1-L5), which are interrelated by one-electron-transfer
steps. The central o-phenylenediamine moiety in H4(L1) can be
oxidized, yielding the iminosemiquinone, H3(L2), and then the
diiminoquinone, H2(L3). Finally, each of the two phenols can
be successively oxidized by one electron, yielding H(L4) and
1
L5. Note that H3(L2) and H(L4) are radicals with an S ) /2
ground state, whereas the species H4(L1), H2(L3), and L5 are
diamagnetic. As we will show below, these five oxidation levels
of the tetradentate ligands (L1)4-, (L2)3-, (L3)2-, (L4)-, and (L5)0
are accessible in their copper(II) and zinc(II) complexes (Chart
2).
The reaction of H4(L1) with [CuI(NCCH3)4](ClO4) or ZnII-
(BF4)2‚2H2O in dry methanol in the presence of air yields
microcrystalline precipitates of green [CuII(L3)] and gray-black
[ZnII(L3)], respectively. Recrystallization of these materials from
acetonitrile afforded single crystals of [MII(L3)]‚CH3CN (MII
) Cu, Zn) suitable for X-ray crystallography. The structure of
the neutral complex in crystals of [Cu(L3)]‚CH3CN is shown
in Figure 1; that of the zinc complex is very similar and not
shown. Table 1 gives selected bond distances and angles for
both complexes.
The respective divalent metal ion in [M(L3)] (M ) Cu, Zn)
is coordinated to the tetradentate ligand (L3)2-; the resulting
N2O2M coordination polyhedron is nearly square planar. The
While the active form of GO containing the CuII-phenoxyl
moiety has been successfully assembled in a number of low-
molecular-weight structural model complexes,4 only very few
functional models have been reported5-7 to date. Stack’s5a,b
functional models A (Chart 1) catalyze the aerobic oxidation
of activated primary alcohols according to eq 1 but not that of
ethanol or methanol. We have reported a model system6 where
a dinuclear CuII-phenoxyl complex B catalyzes the oxidation
of primary alcohols (including ethanol but not methanol). In
addition, B catalyzes also the aerobic oxidation of secondary
alcohols, yielding ketones (eq 2) or pinacols (eq 3).
(5) (a) Wang, Y.; Stack, T. D. P. J. Am. Chem. Soc. 1996, 118, 13097.
(b) Wang, Y.; DuBois, J. L.; Hedman, B.; Hodgson, K. O.; Stack, T. D. P.
Science 1998, 279, 537. (c) Saint-Aman, E.; Menage, S.; Pierre, J.-L.;
Degrancq, E.; Gellou, G. New J. Chem. 1998, 393.
(6) Chaudhuri, P.; Hess, M.; Flo¨rke, U.; Wieghardt, K. Angew. Chem.,
Int. Ed. 1998, 37, 2217.
(7) Chaudhuri, P.; Hess, M.; Weyhermu¨ller, T.; Wieghardt, K. Angew.
Chem., Int. Ed. 1999, 38, 1095.
(8) (a) Kitajima, N.; Whang, K.; Moro-oka, Y.; Uchida, A.; Sasada, Y.
J. Chem. Soc., Chem. Commun. 1986, 1504. (b) Marko, I. E.; Giles, P. R.;
Tsukazaki, M.; Brown, S. M.; Urch, C. J. Science 1996, 274, 2044. (c)
Marko, I. E.; Tsukazaki, M.; Giles, P. R.; Brown, S. M.; Urch, C. J. Angew.
Chem., Int. Ed. Engl. 1997, 36, 2208.

Aerobic Oxidation of Primary Alcohols
J. Am. Chem. Soc., Vol. 121, No. 41, 1999 9601
Scheme 2. Oxidation Levels, Spin States, and Labels of the
Ligand (Only One Resonance Structure Is Given in Each
Case)
Figure 1. Structure of the neutral molecule in crystals of [CuII(L3)]‚CH3-
CN. The structure of [Zn(L3)]‚CH3CN is similar and is not shown.
Table 1. Selected Bond Distances (Å) and Angles (deg)
[Cu(L3)]‚CH3CN
[Zn(L3)]‚CH3CN
M-O1
1.918(2)
1.942(2)
1.954(2)
1.922(2)
1.342(3)
1.373(3)
1.335(3)
1.368(3)
1.433(3)
1.358(3)
1.425(3)
1.362(3)
1.425(3)
1.307(3)
1.318(3)
1.922(2)
1.943(2)
1.955(2)
1.923(2)
1.345(3)
1.370(3)
1.328(3)
1.377(3)
1.436(3)
1.365(3)
1.431(3)
1.362(3)
1.432(3)
1.311(3)
1.320(3)
M-O16
M-N7
M-N14
N7-C8
N7-C6
N14-C13
N14-C15
C8-C9
C9-C10
C10-C11
C11-C12
C12-C13
O1-C1
O16-C16
N7-M-N14
N14-M-O16
O16-M-O1
O1-M-N7
83.17(7)
84.12(7)
109.59(6)
84.65(7)
134.8(2)
112.1(1)
113.0(1)
134.1(2)
114.1(1)
111.8(1)
83.05(8)
84.31(7)
109.42(6)
84.74(7)
135.0(2)
112.1(2)
112.8(1)
133.9(2)
114.6(1)
111.4(1)
C6-N7-C8
C6-N7-M
C8-N7-M
Chart 2. Complexes Prepared, with Their Labels and Their
Electronic Ground States
C13-N14-C15
C13-N14-M
C15-N14-M
Correspondingly, the imino CdN bonds, at 1.34 ( 0.01 Å, are
shorter than the C-N bonds to the phenolates, at 1.37 ( 0.01
Å.
From temperature-dependent magnetic susceptibility mea-
surements (2-290 K), it has been established that [CuII(L3)] is
paramagnetic and [Zn(L3)] is diamagnetic. The copper complex
exhibits a temperature-independent magnetic moment of 1.8 µB
(g ) 2.21) typical for a copper(II) ion (d9).
Reaction of the neutral complexes [M(L3)] (M ) Cu, Zn) in
dry CH2Cl2 with ferroceniumhexafluorophosphate, [Fc]PF6 (1:
1), produces a violet and a green precipitate of [CuII(L4)]PF6
and [ZnII(L4)]PF6, respectively, as depicted in eq 4.
[M(L3)] + [Fc]PF6 f [M(L4)]PF6 + Fc
M ) CuII, ZnII
(4)
square planar Zn(II) complex in a neutral N2O2 coordination is
unprecedented in the Cambridge Structural Database. The
structure determinations unambiguously show that, in both
structures, the oxidation level of the ligand is (L3)2-, containing
a central diiminoquinone moiety and two phenolates. Thus, the
distances C9-C10 and C11-C12, at 1.36 ( 0.01 Å, are
significantly shorter than the distances C8-C9, C10-C11, and
C12-C13, at 1.43 ( 0.01 Å, and C8-C13, at 1.47 ( 0.01 Å.
These complexes are the one-electron oxidized forms of [M(L3)]
(M ) Cu, Zn), respectively. The ligand (L4)- is a paramagnetic
1
organic radical with an S ) /2 ground state. Temperature-
dependent magnetic susceptibility measurements (3-290 K)
reveal a temperature-independent magnetic moment of 1.8 µB
(g ) 2.002) for [Zn(L4)]PF6 containing a diamagnetic ZnII ion

9602 J. Am. Chem. Soc., Vol. 121, No. 41, 1999
Chaudhuri et al.
ature-independent magnetic moment of 1.88 µB (g ) 2.18) has
been established for [CuII(L5)](ClO4)2, indicative of a paramag-
netic CuII ion and diamagnetic L5. Complex [Zn(L5)](ClO4)2 is
diamagnetic.
The reaction of the ligand H4L1 with CuII(ClO4)2‚6H2O or
Zn(ClO4)2‚6H2O (1:1) in dry methanol under an argon atmo-
sphere with 2 equiv of NEt3 yielded brown microcrystals of
[CuII(H2L1)]0 or [ZnII(H2L1)]0. Both compounds are air sensitive
and must be kept under anaerobic conditions. [CuII(H2L1)]0 is
paramagnetic, with an effective magnetic moment of 1.69-1.82
µB in the range 20-300 K and g ) 2.21, whereas [ZnII(H2L1)]0
is diamagnetic.
The solution structure of [ZnII(H2L1)] in dry CDCl3 has been
established by NMR spectroscopy. The 1H NMR spectrum
displays two singlets for the four tert-butyl and three signals in
the aromatic region, with intensities 2H:4H:2H. In addition, one
broad signal at 5.2 ppm (2H) is observed. The relative intensity
of a broad signal at 5.9 ppm (<1 H) is sample dependent and
may correspond to a small impurity of [Zn(H2L1)(H2O)]. In a
2D 1H-15N hetero-multiquantum coherence (HMQC) spectrum,
the signal at δ ) 5.2 shows a cross-peak to a nitrogen at δ )
-315 ppm. The delay for polarization transfer between 1H and
15N was set to 5.5 ms, which corresponds to a coupling constant
of 90 Hz. The presence of this signal proves that this proton is
bound to a nitrogen and not a phenol oxygen. Therefore, we
propose the following structure for the two [MII(H2L1)] com-
plexes. We have not found evidence for species containing
coordinated phenol ligands in solution.
Figure 2. Temperature dependence of the magnetic moment of solid
[CuII(L4)]PF6. The solid line represents a fit using parameters described
in the text.
and a paramagnetic (L4)- ligand. In contrast, [CuII(L4)]PF6 is a
compound containing a CuII(d9) ion and a coordinated radical
(L4)-. Therefore, spin coupling is expected. Figure 2 shows the
temperature dependence of the effective magnetic moment of
solid [CuII(L4)]PF6. At 290 K, the effective magnetic moment
is 2.4 µB, which decreases with decreasing temperature to 0.45
µB at 5 K. We have modeled this temperature dependence by
4
using the usual spin Hamiltonian H ) -2JSL ‚SCu, where J is
the coupling constant and SL ) SCu ) 1/2, with the parameters
4
J ) -48.5 cm-1, gL ) 2.002, and gCu ) 2.23 and 6% of a
4
1
paramagnetic impurity (S ) /2) corresponding to the reduced
form [CuII(L3)]. An antiferromagnetic coupling in [CuII(L4)]-
PF6 generates an S ) 0 ground statesbut see the EPR results
below. This coupling is predominantly inter- rather than
intramolecular in nature, and the coupling constant J does not
represent intramolecular spin exchange coupling between the
organic radical and the CuII ion.
When a blue solution of [Zn(L3)] in dry CH2Cl2 was irradiated
at 20 °C with UV light (Hg immersion lamp), a color change
to bright red was observed within 1.5 h. Addition of [TBA]-
ClO4 initiated the precipitation of diamagnetic red [ZnII(L5)]-
(ClO4)2. Extraction of the residual solution with water allows
the quantitative determination of generated chloride ions ac-
cording to eq 5. 1,2-Dichloroethane has been determined as the
(i) Electro- and Spectroelectrochemistry. Figure 3 displays
the cyclic voltammograms of [M(L3)] (M ) Cu, Zn) in CH2-
Cl2 containing 0.10 M tetra-n-butylammonium hexafluorophos-
phate as supporting electrolyte. Both CVs are quite similar: they
display two successive one-electron reduction and two succes-
sive one-electron oxidation waves, which we assign as shown
in Scheme 3. In each case, the four redox waves involve ligand-
rather than metal-centered one-electron processes. Thus, in the
potential range from +1.5 to -2.0 V vs Fc+/Fc, the copper(II)
ion is not reduced to CuI, since the CV of the corresponding
zinc species recorded in the same potential range is nearly
identical.
Controlled potential coulometry established that the four redox
waves are one-electron processes and that all five redox states
of [Zn(L3)] are stable on the time scale of such an experiment
at ambient temperature (20 min), whereas for [CuII(L3)] the most
reduced form, [CuII(L1)]2-, is not stable in solution.
[ZnII(L3)] + 2CH2Cl2
9
hν8 [ZnII(L5)]2+ + ClH2CCH2Cl +
2Cl- (5)
reduction product by GC/MS and 1H NMR spectroscopy. Thus,
the solvent CH2Cl2 is the effective scavenger of the photo-
chemically generated electrons. Complex [Zn(L5)](ClO4)2 rep-
resents the two-electron oxidized form of [Zn(L3)]. Similarly,
[CuII(L3)] is also photosensitive, but we have only been able to
isolate a solid material consisting of a mixture of [CuII(L4)]X
and [CuII(L5)]X2, where X is PF6- or ClO4-. A pure sample of
red [CuII(L5)](ClO4)2 has been obtained from the reaction of
[CuII(L3)] with 2 equiv of the strong one-electron oxidant [NiIII-
9
(tacn)2](ClO4)3 in CH2Cl2 solution, where tacn is the ligand
Figures 4 and 5 display the electronic spectra of the parent
species [M(L3)] (M ) Cu and Zn, respectively) and their
electrochemically oxidized and reduced forms. The spectral
changes within each series of one-electron-transfer steps are
quite dramatic. The color of a complex in a given oxidation
level is indicated in Scheme 3. Table 2 gives the electronic
spectra of all electrochemically generated species, and Figure
6 displays the spectra of [CuII(H2L1)] and [ZnII(H2L1)] in CH2-
Cl2 solution.
1,4,7-triazacyclononane (eq 6). From temperature-dependent
[CuII(L3)]+ 2[NiIII(tacn)2](ClO4)3 f [CuII(L5)](ClO4)2 +
2[NiII(tacn)2](ClO4)2 (6)
susceptibility measurements in the range 3-290 K, a temper-
(9) Wieghardt, K.; Walz, W.; Nuber, B.; Weiss, J.; Ozarowski, A.;
Stratemeier, H.; Reinen, D. Inorg. Chem. 1986, 25, 1650.

Aerobic Oxidation of Primary Alcohols
J. Am. Chem. Soc., Vol. 121, No. 41, 1999 9603
Figure 3. Cyclic voltammograms of [CuII(L3)] (top) and [Zn(L3)]
(bottom) in CH2Cl2 at 298 K (0.10 M [TBA]PF6 supporting electrolyte;
glassy carbon working electrode; scan rate 100 mV s-1). The potentials
are given in volts vs the Fc+/Fc couple.
Figure 4. Electronic spectra of electrochemically generated oxidized
(top) and reduced (bottom) forms of [CuII(L3)] in CH2Cl2 (0.10 M
[TBA]PF6).
Scheme 3. Summary of the Electrochemistry of [M(L3)] (M
) Cu, Zn) in CH2Cl2 (0.10 M [TBA]PF6)a
a Redox potentials are referenced in volts vs the ferrocenium/
ferrocene couple; the indicated colors of solutions are observed after
coulometry at the appropriate fixed potential.
(ii) EPR Spectroscopy. X-band EPR spectra of [ZnII(L4)]+
and [ZnII(L2)]- in CH2Cl2 solution (0.10 M [TBA]PF6), gener-
ated electrochemically from [Zn(L3)] by one-electron oxidation
and reduction, respectively, have been recorded at 298 K. The
spectra and their simulations are displayed in Figure 7. Both
species exhibit typical S ) 1/2 signals with hyperfine structure.
The spectrum of the monoanion [ZnII(L2)]- displays a signal at
g ) 2.0045 with hyperfine splitting from only one nitrogen atom
and one proton, in excellent agreement with the resonance
structure for an iminosemiquinone moiety in (L2)3- (Scheme
2). In contrast, the spectrum of [ZnII(L4)]+ shows a signal at g
) 2.0045 with hyperfine splittings from two nitrogen atoms
and four different protons, indicating delocalization of the
unpaired electron over the iminoquinone and both phenolate
rings in (L4)-. These spectra prove that the redox chemistry of
Figure 5. Electronic spectra of electrochemically generated oxidized
(top) and reduced (bottom) forms of [Zn(L3)] in CH2Cl2 (0.10 M [TBA]-
PF6).
[ZnII(L3)] is ligand-based. Both [Zn(L5)]2+ and [Zn(L1)]2- are
diamagnetic species.
1-4
Since the redox potentials E1/2
(Scheme 3) for the ZnII
and CuII complexes are very similar, we conclude that the redox
chemistry of [CuII(L3)] is also ligand-based, giving rise to
paramagnetic SCu ) /2 ground states in [CuII(L3)] and [CuII-
1
(L5)](ClO4)2 and [CuII(L1)]2- or [CuII(L1H2)], but St ) 0 ground
states for [CuII(L4)]PF6 and [Cu(L2)]-. The latter are attained

9604 J. Am. Chem. Soc., Vol. 121, No. 41, 1999
Chaudhuri et al.
Table 2. Electronic Spectra of Complexes (300-1600 nm,
CH2Cl2)
λ
max, nm (ꢀ, L mol-1 cm-1
)
[CuII(L5)](ClO4)2 311 (1.0 × 104), 350 sh, 455 (9.1 × 103), 976
(1.25 × 103)
[CuII(L4)]PF6
[CuII(L3)]
320 (5.1 × 103), 478 (6.4 × 103), 981 (1.2 × 103)
315 (9.0 × 103), 356 (8.5 × 103), 453 (2.5 × 103),
483 sh (2.2 × 103), 575 (1.7 × 103), 747
(1.8 × 103), 1256 (5.6 × 103)
[CuII(L2)]- a
)
330 (10.8 × 103), 400 sh, 607 (1.9 × 103), 1073
(2.5 × 103)
[CuII(L1)]2-
334 (1.2 × 103), 369 (855), 600 (170), 1053 (214)
453 (2.5 × 103), 540 sh, 700 sh, 906 (2.5 × 103)
411 (470), 651 (60)
[CuII(L1H2)]b)
[CuI(L1H2)]- b
)
[Zn(L5)](ClO4)2 326 (2.1 × 103), 368 (1.5 × 103), 392 (2.9 × 103),
414 (2.8 × 103), 500 sh, 545 (8.5 × 103), 700 sh
[Zn(L4)]PF6
[Zn(L3)]
385 (4.2 × 103), 411 (3.8 × 103), 578 (2.4 × 103),
810 sh
405 (3.0 × 103), 506 (4.4 × 103), 798 (12.2 × 103)
320 (9.0 × 103), 401 (4.0 × 103), 873 (6.8 × 103)
700 (∼60)
[Zn(L2)]- a
)
)
[Zn(L1)]2- a
[Zn(H2L1)]
454 (5.4 × 103), 800 sh
a Generated by controlled potential coulometry in CH2Cl2 (0.10 M
[TBA]PF6). b Generated in THF solution; [L1H4] ) 10-3 M, [CuCl] or
[CuII(ClO4)2‚6H2O] ) 10-3 M, [NEt3] ) 2 × 10-3 M.
Figure 6. Electronic spectra of [MII(H2L1)] (M ) CuII, ZnII) in CH2-
Cl2.
via intramolecular antiferromagnetic coupling of an unpaired
2
2
electron in a dx -y orbital and a ligand-based unpaired π-elec-
tron.
Complexes[CuII(L3)], [CuII(L5)](ClO4)2, and [CuII(H2L1)]
display typical EPR spectra of square planar CuII complexes
with copper hyperfine structure. From simulations, we obtained
the parameters given in Table 3. The spectra and simulations
are shown in Figures S1, S5, and S2 in the Supporting
Information, respectively.
Interestingly, [Cu(L2)]- in CH2Cl2 is EPR silent at low
temperatures (<70 K) but shows a broad unresolved spectrum
of an S ) 1 system at temperatures >80 K. We take this as
experimental evidence that this species has a diamagnetic ground
state where the ligand radical (L2)3- (S ) 1/2) is antiferromag-
Figure 7. X-band EPR spectra of electrochemically generated [Zn-
(L4)]+ (top) and [Zn(L2)]- (bottom) in CH2Cl2 (0.10 M [TBA]PF6) at
298 K and their simulations. Experimental conditions: [complex],
∼10-4 M; frequency, 9.4501 GHz; microwave power, 10 mW;
modulation amplitude, 0.05 mT; gain, 3200. The hyperfine coupling
constants are given in millitesla.
1
netically coupled to a CuII ion (S ) /2).
Similarly, the spectrum of [CuII(L4)]PF6 in CH2Cl2 shows
signals at geff ≈ 2.0 and a typical half-field signal of an S ) 1
system at geff ≈ 4.0 at 50 K (Figure S3, Supporting Information).
The latter signal disappears upon decreasing the temperature.
From its temperature dependence, an antiferromagnetic coupling
constant of J ) -7 ( 3 cm-1 was calculated (Figure S4,
Supporting Information) by using the spin Hamiltonian H )
-2JSrad‚SCu (Srad ) SCu ) 1/2). This intramolecular coupling of
-7 cm-1 in solution is significantly smaller than the value
obtained from temperature-dependent magnetic susceptibility
measurements on a solid sample of [CuII(L4)]PF6 (see above).
Table 3. EPR Spectra of CuII Complexes in CH2Cl2 Solution
Ax ) Ay,
Az,
complex
T, K
gx
gy
gz
×104 cm-1 ×104 cm-1
[CuII(H2L1)]
[CuII(L3)]0
10 2.059 2.067 2.278
10 2.062 2.062 2.278
4
11
10
32
183
155
156
[CuII(L5)](ClO4)2
2.044 2.076 2.224
This indicates that, in addition to a relatively small intramo-
lecular exchange coupling, there also exists a relatively strong
intermolecular interaction in the solid state.

Aerobic Oxidation of Primary Alcohols
J. Am. Chem. Soc., Vol. 121, No. 41, 1999 9605
Single Turnover of [CuII(L4)]+ and [ZnII(L4)]+ with
Primary Alcohols. When a solution of [CuII(L4)]PF6 and NEt3
(1:2) in dry CH2Cl2 was added to an excess of a primary alcohol
such as methanol, ethanol, or benzyl alcohol under anaerobic
conditions (Ar) at 22 ( 1 °C, the original violet color changed
to deep blue within minutes. The final UV/vis spectrum of this
solution corresponds to that of [CuII(L2)]- generated electro-
Table 4. Kinetic Data of Anaerobic Single-Turnover Reactions at
22 ( 1 °C
KH/ kH/
KD kD
oxidant
reductant condnsa K, M-1
k, s-1
[CuII(L4)]+ CH3OH
1
1
2
2
3
3
79 ( 15 (1.2 ( 0.6)
× 10-5
1.01 54
[CuII(L4)]+ CD3OH
78 ( 5 (2.9 ( 0.05)
× 10-7
1
chemically from [CuII(L3)] in CH2Cl2. H NMR/GC analysis
[CuII(L4)]+ CH3CH2OH
[CuII(L4)]+ CH3CD2OH
[Zn(L4)]+ CH3CH2OH
[Zn(L4)]+ CH3CD2OH
20 ( 6 (2.0 ( 0.3)
× 10-5
of the resulting solution indicates the formation of formaldehyde,
acetaldehyde, and benzaldehyde, respectively. Spectrophoto-
metric quantitative determination of these aldehydes10 indicates
a stoichiometry as in eq 7 with >95% yield of aldehyde and
[CuII(L2)]-, respectively.
0.95 14
1.33 15
21 ( 2 (1.4 ( 0.6)
× 10-6
12 ( 3 (1.6 ( 0.3)
× 10-6
9 ( 2 (1.1 ( 0.2)
× 10-7
+2NEt3
R-CH2OH + [CuII(L4)]+
8 R-CHO +
a Reaction conditions: (1) solvent CH2Cl2, [CuII(L4)+] ) 3.5 × 10-4
M, [methanol] ) 2.5 × 10-3-2.0 × 10-2 M, [NEt3] ) 7.1 × 10-4 M;
(2) solvent CH2Cl2 containing 0.10 M [TBA]PF6, electrochemically
generated [Cu(L4)]+ (from [Cu(L3)], 3.5 × 10-4 M), [ethanol] ) 8.0
× 10-3-8.0 × 10-2 M, [NEt3] ) 7.1 × 10-4 M; (3) solvent CH2Cl2,
[Zn(L4)+] ) 3.5 × 10-4 M, [ethanol] ) 5 × 10-3-0.25 M, [NEt3]
)7.1 × 10-4 M. All reactions were run under an Ar blanketing
atmosphere.
-2[HNEt3]+
[CuII(L2)]- (7)
R ) H, CH3, phenyl
Thus, complex [CuII(L4)]+ is capable of oxidizing primary
alcohols to the corresponding aldehyde and forming the two-
electron reduced form [CuII(L2)]-. No reaction was observed
when using secondary alcohols (2-propanol or diphenylcarbinol)
using the same conditions.
Similarly, the zinc-containing monocationic complex [Zn-
(L4)]+ undergoes the same reaction (eq 8) as was established
by following the reaction spectrophotometrically and with GC
product analysis.
contained the same overall copper(II) concentration. We then
prepared a fourth methanol solution in which we combined two
aliquots of solutions containing (a) [CuII(L4)]+ and (b) [CuII(L2)]-,
and we measured the absorbance at 450 nm. If the latter solution
contains equal amounts of unreacted [Cu(L4)]+ and [Cu(L2)]-,
then the observed molar extinction coefficient, ꢀ450, is expected
to be 3.6 × 103 L mol-1 cm-1; if, on the other hand,
comproportionation has taken place, ꢀ450 is expected to be 1.6
× 103 L mol-1 cm-1. The experimental value, ꢀ450 ) 3.5 ×
103 L mol-1 cm-1, confirms that, in methanol solution, no
comproportionation between [Cu(L4)]+ and [Cu(L2)]- occurs.
The kinetics of the reactions of [M(L4)]+ (M ) Cu, Zn) with
methanol and ethanol (eqs 7 and 8, respectively) have been
measured spectrophotometrically in CH2Cl2 solution at 22 ( 1
°C under an argon blanketing atmosphere. Two equivalents of
NEt3 (7.1 × 10-4 M) had been added in each experiment as a
proton scavenger. The spectral changes at 454 nm for 2 and at
578 nm for 5 have been measured as a function of time using
pseudo-first-order conditions with an excess alcohol (0.005-
0.25 M) and [complex] ) 3.5 × 10-4 M. First-order rate
constants, kobs, plotted vs the alcohol concentration displayed
saturation behavior as in eq 10. Numerical values for the
+2NEt3
R-CH2OH + [ZnII(L4)]+
8 R-CHO + [ZnII(L2)]-
-2[HNEt3]+
(8)
R ) H, CH3, phenyl
The final UV/vis spectrum of the solution is identical to that of
electrochemically generated [Zn(L2)]- (from [Zn(L3)]0 via
coulometric one-electron reduction).
Interestingly, neither [CuII(L3)] nor [ZnII(L3)] reacts with the
above alcohols. Both species have not been detected as
intermediates or products in the reaction of [CuII(L4)]+ and [Zn-
(L4)]+ with primary alcohols. In contrast, upon mixing of two
aliquots of CH2Cl2 solutions containing only electrochemically
generated (a) [CuII(L4)]+ (or [ZnII(L4)]+) and (b) [CuII(L2)]-
(or [ZnII(L2)]-), the spectrum of [CuII(L3)]0 (or [ZnII(L3)]0) is
generated within mixing time according to eq 9 via compro-
portionation.
kK[alcohol]
kobs
)
(10)
1 + K[alcohol]
equilibrium constants K (in M-1) and the rate constant k (in
s-1) were evaluated from a least-squares fit of the data to eq 10
and are summarized in Table 4.
The simplest mechanism in accord with the above kinetic
data involves reversible binding of one molecule of alcohol to
the complex [M(L4)]+ (M ) Cu, Zn) and irreversible oxidation
of it in the rate-determining step (eqs 11 and 12).
[MII(L4)]+ + [MII(L2)]- f 2[MII(L3)]0
MII ) Cu or Zn
(9)
We take the fact that the [M(L3)] species are not observed in
the slow reduction of [M(L4)]+ by an alcohol as an indication
that, under our experimental conditionssusing a large excess
of alcohol over the [M(L4)]+ speciessa five-coordinate species,
[M(HL4)(OR)]+, prevails which does not undergo a one-electron
reduction by the product [M(L2)]-. We have tested this
assumption by using the following experiment. The absorption
coefficients of a solution containing [CuII(L4)]+ and, indepen-
dently, of solutions containing [CuII(L3)] and of [Cu(L2)]- in
methanol at 450 nm were measured. Each of the three solutions
K
[M(L4)]+ + RCH2OH y
\
z [M(L4)(OCH2R)] + H+ (11)
[M(L4)(OR)]
9
k8 [M(L2)]- + RCHO + H+
(12)
The kinetics of the same reactions using the selectively
deuterated substrates CD3OH and CH3CD2OH have also been
measured under the same reaction conditions as for the unlabeled
alcohols. Significantly, the ratio KH/KD is in the range 0.95-
1.33, which implies that the binding of the substrate does not
(10) Kaka´c, B.; Vejdelek, Z. J. Handbuch der photometrischen Analyse
organischer Verbindungen; Verlag Chemie, Weinheim, 1974; Vol. 1, pp
257 and 259.

9606 J. Am. Chem. Soc., Vol. 121, No. 41, 1999
Chaudhuri et al.
Scheme 4
exhibit a marked isotope effect. In contrast, the kH/kD ratios of
54 for the reaction of methanol with [CuII(L4)]+ and of 14 ( 1
for the reactions of [M(L4)]+ (M ) Cu, Zn) with ethanol indicate
a significant large kinetic isotope effect (KIE). We take this as
clear evidence that H-abstraction from the R-carbon atom of
the coordinated alcohol is the rate-determining step. The
resulting coordinated ketyl radical anion is known to be a strong
one-electron reductant11 which is converted to the aldehyde via
an intramolecular one-electron-transfer step. In the present cases,
this intramolecular electron-transfer step reduces the ligand from
the (L3)2- to the (L2)3- oxidation level.
Interestingly, the stoichiometric reaction of [CuII(L4)](PF6)
(3.1 × 10-4 M) with 2-chloroethanol (3.1 × 10-3 M) under an
argon atmosphere in the presence of triethylamine (6.3 × 10-4
M) in dry THF at 22 °C produces within 20 min quantitatively
the one-electron reduced complex [CuII(L3)], succinic dialdehyde
(95%), and free chloride ions (95%) according to the reaction
sequence shown in Scheme 4. In contrast, as shown above, the
same reaction with the substrate ethanol produces [CuII(L2)]-
and acetaldehyde. It is well established that the ketyl radical
anion -O-C˙ H-CH2Cl rearranges intramolecularly very rapidly
(k > 109 s-1), with formation of free chloride ions and the
C-centered radical C˙ H2CHO,3a,12 which is a poor ligand and
dimerizes in solution with formation of succinic dialdehyde.
We have not been able to identify any chloroacetaldehyde
(<1%) (or [Cu(L2)]-) as product in the above reaction. This
proves that H atom abstraction is the rate-determining step of
alcohol oxidation by [M(L4)]+ (M ) CuII, ZnII) and not hydride
transfer, which would necessitate the products chloroacetalde-
hyde and [CuII(L2)]-. Furthermore, the quantitative formation
of succinic dialdehyde and chloride ions indicates that the
alternative intramolecular electron transfer from the coordinated
ketyl radical anion to the CuII(L3) unit is significantly slower
(at least 2 orders of magnitude) than the above intramolecular
rearrangement of the radical with expulsion of Cl- (i.e., ket ,
krearr ) 109 s-1). The reaction of [Zn(L4)]PF6 with 2-chloroet-
hanol also yields quantitatively one-electron reduced [Zn(L3)],
succinic dialdehyde, and chloride ions.
Figure 8. Plot of the turnover versus time of the catalyses at 22 ( 1
°C in THF. Top: [Cu(L4)] ) 2.5 × 10-4 M, [NEt3] ) 5.0 × 10-4 M,
[ethanol] ) 2.5 M in the presence of air (1 bar). Bottom: [Zn(L4)+] )
2.5 × 10-4 M, [NEt3] ) 5.0 × 10-4 M, [ethanol] ) 2.5 M (9); and
[Zn(L3)] ) 2.5 × 10-4 M, [NEt3] ) 5.0 × 10-4 M, [ethanol] ) 2.5 M
(O) in the presence of air. The turnover is defined as the ratio
[acetaldehyde]/[catalyst] or (indistinguishably) [H2O2]/[catalyst] at a
given point in time.
Reaction of [CuII(L2)]- and [Zn(L2)]- with O2. When a
solution containing electrochemically generated [CuII(L2)]- (3.5
× 10-4 M) and acetic acid (7.1 × 10-4 M) in CH2Cl2 (0.1 M
[TBA]PF6) was exposed to dioxygen, the color changed within
2-3 min from blue to violet. The electronic absorption spectrum
of this solution corresponds to that of [CuII(L4)]+ in CH2Cl2
(0.1 M [TBA]PF6). In addition, 1 equiv of H2O2 has been
detected. Thus, the rapid reaction of [CuII(L2)]- with O2
quantitatively yields H2O2 and [Cu(L4)]+ (eq 13) in the presence
of protons (acetic acid). Similar experiments with [ZnII(L2)]-
[CuII(L2)]- + O2 + 2H+ f [CuII(L4)]+ + H2O2 (13)
and O2 produced also H2O2 and [ZnII(L4)]+ (eq 14). Reactions
[ZnII(L2)]- + O2 + 2H+ f [ZnII(L4)]+ + H2O2 (14)
13 and 14 represent formally a two-electron oxidation of the
ligand (L2)3- to (L4)-.
Homogeneous Catalysis. The stoichiometric reactions equa-
tions (eqs 7, 8 and eqs 13, 14) taken together immediately imply
that, in principle, the oxidation of primary alcohols by dioxygen
should be catalyzed by the complexes [M(L4)]+ (M ) Cu, Zn).
This is, indeed, the case.
Figure 8a shows the result of an experiment where the
formation of acetaldehyde and H2O2 from ethanol (2.5 M) and
air (1 bar) in THF in the presence of the catalyst [Cu(L4)]+
(2.5 × 10-4 M) was monitored at 22 °C as a function of time.
The catalysis stopped after ∼45 h. For the first 40 h, the turnover
increases linearly with time, where the turnover is defined as
the ratio [aldehyde]/[catalyst] or (indistinguishably) [H2O2]/
[catalyst] at a given point in time. During this time, the color
of the solution is greenish, and its electronic spectrum resembles
that of [Cu(L4)]+. After ∼40 h, the color changed to yellow, at
which point the reaction stopped rather abruptly; ∼50%
(11) Wang, W.-F.; Schuchmann, M. N.; Bachler, V.; Schuchmann, H.-
P.; von Sonntag, C. J. Phys. Chem. 1996, 100, 15843 and references therein.
(12) (a) Tanner, D. D.; Chen, J. J.; Chen, L.; Luelo, C. J. Am. Chem.
Soc. 1991, 113, 8074. (b) Tanner, D. D.; Yang, C.-M. J. Org. Chem. 1993,
58, 5907. (c) Tanner, D. D.; Xie, G.-J.; Hooz, J.; Yang, C.-M. J. Org. Chem.
1993, 58, 7138.

Aerobic Oxidation of Primary Alcohols
J. Am. Chem. Soc., Vol. 121, No. 41, 1999 9607
Table 5. Kinetic Data of Catalytic Reactions at 22 + 1 °C
Interestingly, the alcohol binding constants, KH and KD,
measured under turnover conditions are of the same order of
magnitude as those determined for the stoichiometric reactions
(eqs 7, 8). The difference is mainly due to the fact that the
stoichiometric reactions were run in CH2Cl2 (a noncoordinating
solvent), whereas the catalysis was performed in THF (weakly
coordinating). The ratio KH/KD is in the range 0.92-1.5,
indicating the lack of a significant isotope effect. In contrast,
the ratio kH/kD under turnover conditions is 36 and 10 for
methanol and ethanol substrates, respectively, indicating a
KH/ kH/
catalyst
alcohol
condnsa
1
K, M-1
12 ( 2
k, s-1
KD kD
[Cu(L4)]+ CH3OH
(5.0 ( 0.3)
× 10-6
0.92 36
0.93 10
1.5 11
CD3OH
1
1
1
2
2
13 ( 3
(1.4 ( 0.2)
× 10-7
[Cu(L4)]+ CH3CH2OH
CH3CD2OH
0.5 ( 0.1
(1.6 ( 0.09)
× 10-5
0.54 ( 0.09 (1.6 ( 0.2)
× 10-6
[Zn(L4)]+ CH3CH2OH
CH3CD2OH
0.6 ( 0.2
0.4 ( 0.1
(6.8 ( 1.3)
× 10-7
cat
cat
significant KIE. The numerical values for kH and kD are
very similar to those found for the stoichiometric reactions (kH,
kD). This implies that reoxidation of the reduced catalysts by
oxygen is fast compared to the oxidation of alcohols.
(6.0 ( 0.1)
× 10-8
a Reaction conditions: (1) [H4L1] ) 2.5 × 10-4 M, [CuICl] ) 2.5
× 10-4 M, [NEt3] ) 10-3 M in tetrahydrofuran, [alcohol] ) 0.01-2.5
M, stirred under 1 bar of air at 22 ( 1 °C; (2) as in (1) but using
Zn(ClO4)2‚6H2O instead of CuICl.
The catalysts can also be assembled in situ from solutions of
the ligand H4L1 and CuCl (1:1) or ZnII(ClO4)2‚6H2O in THF
under anaerobic conditions. Upon exposure to air, the catalyti-
cally active forms [Cu(L4)]+ and [Zn(L4)]+ are generated with
formation of 2 and 1.5 equiv of H2O2, respectively, as depicted
in eqs 16 and 17.
conversion of the substrate ethanol had been achieved. In a
similar experiment, which is shown in Figure 8b, using identical
reaction conditions except that [Zn(L4)]+ was used as catalyst
(2.5 × 10-4 M), again a linear increase of turnovers with time
is observed but only for ∼25 h, after which time the reaction
stopped. Within 25 h, only 170 turnovers were achieved; the
conversion was ∼2%. Thus, the zinc is a far less effective and
stable catalyst than its copper analogue. When the same
experiments as described above were performed using [CuII-
(L3)] and [Zn(L3)] as catalyst, a pronounced kinetic lag phase
was observed; both species are not the catalytically competent
forms of the complexes.
We have measured the kinetics of the catalyses under turnover
conditions by using the initial rate method, where the concentra-
tion of the respective catalyst was varied in the range 2.5 ×
10-4-2.5 × 10-3 M and that of the alcohol (methanol and
ethanol) in the range 0.01-2.5 M and [NEt3] ) 2[catalyst]. The
formation of the aldehyde and, independently, of H2O2 was
monitored spectrophotometrically as a function of time (see
Experimental Section for details). The rate of formation of both
products is identical. This implies that the catalysts do not
catalyze the disproportionation of H2O2. A rate law (eq 15) was
thus established.
+O2
H4L1 + CuCl f [CuI(L1H4)]+ -H O 8 f
2
2
+O2
[CuII(L2H2)]+ -H O 8 [CuII(L4)]+ (16)
2
2
+O2
H4L1 + Zn2+
8 [ZnII(L4)]+ + 1.5H2O2 + H+ (17)
The stoichiometry of these reactions has been determined
spectrophotometrically ([CuII(L4)]+ and [Zn(L4)]+), and the
amount of H2O2 produced has been measured spectrophoto-
metrically after its conversion to the peroxotitanyl species (see
Experimental Section).
Discussion
In this work, we have presented two mononuclear coordina-
tion compounds, [CuII(L4)]PF6 and [Zn(L4)]PF6, which catalyze
the selective aerobic oxidation of primary alcohols, including
methanol, yielding aldehydes and 1 equiv of H2O2. In this sense,
they represent functional models for the enzyme GO. Mecha-
nistically, there are subtle differences due to the fact that [Zn-
(L4)]PF6, containing a redox-inactive central zinc(II) ion, is also
an active catalyst although not quite as efficient as [CuII(L4)]-
PF6.
kK[alcohol]
rate )
[catalyst]
(15)
(
)
1 + K[alcohol]
We have also briefly studied the effect of [NEt3] on the
catalysis.7 In a series of experiments with [ethanol] ) 0.125 M
) constant and [Cu(L4)+] ) 2.5 × 10-4 M ) constant and
variation of [NEt3] ) 5.0 × 10-4-2.0 × 10-3 M, it was found
that, when the ratio [NEt3]/[Cu(L4)+] > 4, the catalysis is
effectively inhibited, and at a ratio ∼5 catalysis does not
proceed. This is a clear indication that the base triethylamine is
a competitive inhibitor which binds more effectively to [Cu-
(L4)]+ than the substrate alcohol. We also note that acetonitrile
is a very effective inhibitor.
In Figure 9, we present a minimal mechanistic scheme for
the catalysis which is in accord with all the structural and kinetic
information assembled above. In the first step, a rapid equilib-
rium between the catalyst and one molecule of alcohol is
established. The alcohol is most probably bound in its anionic
alcoholate form, where the proton from the bound alcohol is
transferred either to the external base NEt3 or to the phenolate
ligand with formation of a coordinated phenol. Such coordinated
phenols have been shown to possess a dissociation constant,
pKa, of ∼5-6 in aqueous solution.4b Coordination of an
alcoholate ligand to [M(L4)]+ (M ) Cu, Zn) results in a five-
coordinate, square-base pyramidal intermediate, where the
axially bound alcoholate is in cis-position relative to the
phenoxyl radical ligand.
Table 5 summarizes numerical values of k and K obtained
from a least-squares fit the kinetic data to eq 15. We have also
investigated the reaction with the selectively deuterium labeled
substrates CD3OH and CH3CD2OH.
The observed first-order dependence of the rate on the
concentration of [Cu(L4)]PF6 or [Zn(L4)]PF6 clearly indicates
that the monomeric cations represent the catalytically active
forms and not a dinuclear species, such as complex B in Chart
1.7 The observed saturation kinetics for the catalyses allows one
to determine the binding constants of the substrate to the catalyst.
In this geometrical arrangement, one hydrogen atom of the
R-carbon atom of the alcoholate ligand is, in principle, capable
of forming a hydrogen bond to the phenoxyl radical ligand,
giving rise to a five-membered chelate ring Cu-O˙ ‚‚‚H-C-O.

9608 J. Am. Chem. Soc., Vol. 121, No. 41, 1999
Chaudhuri et al.
complexes,16,17 a rate-determining hydride transfer from the
R-carbon atom of the alcohol to the oxo metal moiety has been
proposed. In general, a kH/kD of >10 (at 20 °C) has been
interpreted as the occurrence of nuclear tunneling,19 but it is
not possible to distinguish between a hydride and a hydrogen
atom transfer mechanism (i.e., hetero- vs homolytic C-H
scission) by the observed ratio kH/kD. As Roecker and Meyer
have pointed out,16 “the experimental facts pertaining to kH/kD
isotope effects show that they are not amenable to simple
interpretation”.
It is noteworthy in this respect that Whittaker et al.3d have
recently reported a detailed study of the KIEs as probes of the
mechanism of GO. While steady-state kinetics studies have
demonstrated a KIE of 7-8 that was attributed to substrate
oxidation (galactose), these authors have used rapid kinetic
methods to measure the anaerobic reduction of GO substrate
by galactose and found KIEs of 22.5 (4 °C).
Branchaud et al.3a,20 used simple â-haloethanols as mecha-
nistic probes and demonstrated rapid enzyme inactivation via
reduction of the active site to the inactive CuII form. They
concluded that the substrate oxidation step of GO proceeds either
“through a short-lived ketyl radical anion intermediate or through
a closely related concerted E2R mechanism with considerable
ketyl radical anion character in the transition state”.
Figure 9. Proposed mechanism for the catalytic oxidation of primary
alcohols by dioxygen.
In this work, the reduction of the radical complex [CuII(L4)]+
by the substrates â-chloroethanol on one hand and ethanol on
the other, yielding quantitatively [Cu(L3)] and [Cu(L2)]-,
respectively, and as oxidation products succinic dialdehyde and
acetaldehyde, respectively, lends strong support to the notion
that a common coordinated ketyl radical anion is formed as an
intermediate in the rate-determining step of alcohol oxidation
by the catalyst [CuII(L4)]+ (and [Zn(L4)]+).
This has been beautifully demonstrated by Tolman et al.,4a,13
who have shown by X-ray crystallography that the following
intramolecular hydrogen-bonding scheme prevails in a mono-
nuclear phenolato(benzyl alcoholato)copper(II) complex. These
The resulting intermediate, containing an O-coordinated ketyl
radical anion, is proposed to undergo a rapid intramolecular one-
electron transfer with formation of a bound aldehyde, which is
a poor ligand and dissociates. The ultimate electron acceptor
of this step is the o-diiminoquinone unit of the doubly protonated
ligand H2L3, which is reduced to [H2L2]-. Mechanistically, it
is conceivable that, in the case of the copper(II)-containing
catayst, this ion is the primary electron acceptor, which is
thereby reduced to Cu(I) and then intramolecularly reoxidized
to CuII with formation of coordinated [H2L2]-. In the case of
[Zn(L4)]+, this is not possible. We have no experimental
evidence for either pathway, as these steps follow the rate-
determining H-abstraction step.
After the intramolecular oxidation of the ketyl radical anion
and dissociation of the aldehyde, the catalysts are in the reduced
[MII(L2H2)]+ (MII ) Cu, Zn) state, where most likely the two
phenolates are protonated (or 2 equiv of [HNEt3]+ has been
formed). The reduced catalysts react with dioxygen, yielding
H2O2 and [M(L4)]+sthe active catalyst. This reoxidation step
most probably is also broken down into two successive one-
electron-transfer steps, where a coordinated superoxide metal-
(II) intermediate is formed (eq 18). Such a diamagnetic
authors have also shown that electrochemical one-electron
oxidation of the phenolate to the corresponding phenoxyl
triggers an internal redox process whereby benzaldehyde is
produced presumably along with CuI. We suggest that the above
five-membered ring resembles the transition state for the
H-abstraction reaction in our catalytic systems and in Tolman’s
model complex. Clearly, this is the rate-determining step in the
catalyses with [M(L4)]+ (M ) Cu, Zn), since this intramolecular
H-abstraction displays large kinetic isotope effects.
These large KIEs deserve a comment. It is well established
that oxidation of primary alcohols by oxo metal complexes such
- 14
as MnO4
,
RuO4,15 [(bpy)2(py)Ru(O)]2+ 16
, [(trpy)(bpy)Ru-
[MII(L2H2)]+ + O2 f [MII(L3H2)(O2•-)]+ f [MII(L4)]+ +
(O)]2+ 17
,
and, more recently, trans-[RuVI(trpy)/(O)2(L)]2+ (L )
H2O, CH3CN)18 display in some instances very large KIEs (up
H2O2 (18)
to kH/kD ) 61.5), and, at least for the first two ruthenium
intermediate has recently been studied.7 It has been shown that,
in THF solution, the following equilibria (eq 19) exist, where
(13) Halfen, J. A.; Young, V. G., Jr.; Tolman, W. B. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 1687.
(14) (a) Stewart, R. J. Am. Chem. Soc. 1957, 79, 3057. (b) Stewart, R.;
van Linden, R. Discuss. Faraday Soc. 1960, 29, 211.
(15) Lee, D. G.; van den Engh, M. Can. J. Chem. 1972, 50, 2000.
(16) Roecker, L.; Meyer, T. J. J. Am. Chem. Soc. 1987, 109, 746.
(17) Thompson, M. S.; Meyer, T. J. J. Am. Chem. Soc. 1982, 104, 4106.
(18) Lebeau, E. L.; Meyer, T. J. Inorg. Chem. 1999, 38, 2174.
(19) (a) Cha, Y.; Murray, C. J.; Klinman, J. P. Science 1989, 243, 1325.
(b) Bell, R. P. The Tunnel Effect in Chemistry; Chapman and Hall: New
York, 1980.
(20) Wachter, R. M.; Branchaud, B. P. Biochim. Biophys. Acta 1998,
1384, 43.

Aerobic Oxidation of Primary Alcohols
J. Am. Chem. Soc., Vol. 121, No. 41, 1999 9609
H3(Lc) and H2(Lb) represent the tridentate ligand N,N′-bis(2-
hydroxy-3,5-di-tert-butylphenyl)amine and its monooxidized
radical form, respectively.
(21%). Anal. Calcd for C34H44N2O10Cl2Cu: C, 52.68; H, 5.72; N, 3.61;
Cu, 8.20. Found: C, 52.46; H, 5.70; N, 3.60; Cu, 8.29.
[ZnII(L3)]. A solution of Zn(BF4)2‚2H2O (239 mg; 1.0 mmol), the
ligand H4L1 (516 mg, 1.0 mmol), and NEt3 (0.5 mL) in dry methanol
(50 mL) was heated to reflux under an argon atmosphere for 30 min.
After cooling of the yellow solution to 20 °C, it was exposed to air
with stirring, whereupon the color changed to deep green. Within 24 h
at ambient temperature, microcrystalline gray-black crystals formed,
which were collected by filtration. Recrystallization from acetonitrile
yielded single crystals of [Zn(L4)]‚CH3CN. Yield: 320 mg (62%). 1H
NMR (400 Hz, CDCl3): δ 1.19 (s, 18H), 1.21 (s, 18H), 6.41 (m, 2H),
6.69 (m, 2H), 6.72 (b, 2H), 7.08 (d, J ) 1.7 Hz, 2H). The signals at
6.41 and 6.69 ppm form an AA′XX′ system, which was satisfactorily
simulated with the following coupling constants: JAA′ ) 9.8 Hz, JA′X′
) 6.8 Hz, JAX′ ) 1.5 Hz, JAX ) 1.3 Hz. 13C{1H} NMR (100.6 MHz,
CDCl3): δ 29.39, 31.30, 34.33, 35.18, 116.36, 122.65, 126.29, 129.61,
136.02, 140.02. Anal. Calcd for C34H44N2O2Zn: C, 70.64; H, 7.67; N,
4.85; Zn, 11.31. Found: C, 70.45; H, 7.66; N, 4.86; Zn, 11.53.
[CuI(LcH2)(NEt3)] + O2 h [CuII(LcH2)(O2•-)(NEt3)] h
[CuII(Lb)(NEt3)] + H2O2 (19)
Conclusion
We presented in this work two catalysts, [M(L4)]PF6 (M )
Cu, Zn), which effectively catalyze the aerobic oxidation of
primary alcohols, including ethanol and methanol, with forma-
tion of the corresponding aldehydes and H2O2 at ambient
temperature. Up to ∼5 × 103 turnovers in 50 h have been
achieved for the air oxidation of ethanol by [Cu(L4)]+, which
corresponds to a turnover frequency of ∼0.03 s-1. In contrast,
the zinc analogue is less efficient, and the catalyst stability is
inferior (170 turnovers in 24 h corresponds to a turnover
frequency of 0.002 s-1). This reactivity difference reflects the
fact that the intramolecular rate constant for H-abstraction is
faster by a factor of ∼10 for the copper-containing catalyst than
for its zinc analogue. None of the previously reported catalysts5b,6,7
is capable of oxidizing methanol. The catalysts [M(L4)]+ (M )
Cu, Zn) are the most effective and stable ones (with respect to
decomposition during turnover) reported to date, but overall the
process is slow.
[ZnII(L4)](PF6). A solution of [Zn(L3)] (146 mg; 0.25 mmol) and
ferrocenium hexafluorophosphate (83 mg; 0.25 mmol) in dry CH2Cl2
(30 mL) under an Ar atmosphere was stirred at -10 °C for 2 h and
then stored at -80 °C for 12 h. A green microcrystalline precipitate
1
formed, which was collected by filtration. Yield: 63 mg (35%). H
NMR (400 MHz, CDCl3): δ -5.11 (1H), -1.19 (1H), 1.10 (1H), 1.39
(9H), 1.89 (9H), 6.46 (2H), 7.23 (1H), 8.29 (1H), 18.04 (1H), 34.87
(1H), 55.38 (1H), 56.13 (1H). Anal. Calcd for C34H44N2O2PF6Zn: C,
56.47; H, 6.13; N, 3.88; Zn, 9.04. Found: C, 56.28; H, 6.15; N, 3.85;
Zn, 8.89.
[ZnII(L5)](ClO4)2. A solution of [Zn(L3)] (578 mg; 1.0 mmol) in
dry CH2Cl2 (100 mL) was photolyzed for 1.5 h at ambient temperature
with a Hg immersion lamp. A color change of the deep blue solution
to light red was observed during this time. The volume of the reaction
mixture was reduced to 50 mL under reduced pressure at -20 °C, and
[TBA]ClO4 (0.68 g; 2 mmol) was added. After the solution was stored
at -80 °C for 2 d, a red microcrystalline precipitate had formed, which
Experimental Section
N,N′-Bis(3,5-di-tert-butyl-2-hydroxyphenyl)-1,2-phenylenedi-
amine (H4L1). A solution of 3,5-di-tert-butylcatechol (8.9 g; 4.0 mmol),
o-phenylenediamine (2.06 g; 2.0 mmol), and triethylamine (0.4 mL)
in n-heptane (120 mL) was stirred at ambient temperature in an open
vessel for 4 d, after which time a pale yellow precipitate was collected
by filtration and washed with n-pentane. Yield: 4.6 g (44%). 1H NMR
(250 MHz, CDCl3): δ 1.30 (s, 18H), 1.48 (s, 18H), 5.30 (b, 2H), 5.85
(b, 2H), multiplet 6.70 (b), 6.92 (b), 7.18 (b) (8H). 13C{1H} NMR (62.89
MHz, CDCl3): δ 29.69, 31.62, 34.48, 34.87, 117.46, 118.72, 120.41,
122.20, 128.82, 135.47, 142.65, 147.09. ESI MS: calcd for C34H48N2O2
516.8, found 516.3. Anal. Calcd for C34H48N2O2: C, 79.02; H, 9.36;
N, 5.42. Found: C, 78.86; H, 9.40; N, 5.43.
1
was collected by filtration at -20 °C. Yield: 80 mg (10%). H NMR
(400 Hz, CDCl3): δ 0.97 (s, 18H), 1.18 (s, 18H), 6.78 (m, 2H), 7.00
(b, 2H), 7.07 (b, 2H), 7.33 (m, 2H). The signals at 6.78 and 7.33 ppm
form an AA′XX′ system, which was satisfactorily simulated with the
coupling constants: JAA′ ) 9.4 Hz, JA′X′ ) 6.2 Hz, JAX′ ) 1.1 Hz, JAX
) 1.1 Hz. 13C{1H} NMR (62.9 MHz, CDCl3): δ 29.04, 31.08, 34.14,
34.55, 114.97, 123.52, 126.94, 129.90, 131.75, 138.53, 138.86, 144.78,
160.38. Anal. Calcd for C34H44N2O10ZnCl2: C, 52.55; H, 5.70; N, 3.61.
Found: C, 52.36; H, 5.72; N, 3.57.
[CuII(L3)]. An anaerobic solution of the ligand H4L1 (0.52 g; 1.0
mmol), [CuI(NCCH3)4](ClO4) (0.37 g; 1.0 mmol), and NEt3 (0.5 mL)
in dry methanol (50 mL) was heated to reflux under an Ar blanketing
atmosphere for 30 min, after which time the yellow solution was cooled
to 20 °C and exposed to air. From the green solution, green
microcrystals precipitated within 2 h, which were collected by filtration.
Recrystallization of this material from CH3CN produced green single
[CuII(H2L1)]. To degassed solution of H4L1 (517 mg; 1.0 mmol)
and CuII(ClO4)2‚6H2O (370 mg; 1.0 mmol) in dry, freshly distilled CH3-
OH (50 mL) was added NEt3 (280 µL) under an Ar blanketing
atmosphere. After the solution was stirred for 30 min at 20 °C, a brown
precipitate had formed in ∼70% yield. Anal. Calcd for C34H46N2O2-
Cu: C, 70.62; H, 8.02; N, 4.84; Cu, 10.99. Found: C, 70.49; H, 7.95;
N, 4.83; Cu, 10.83. The complex is air-sensitive and was stored under
an Ar atmosphere.
[ZnII(H2L1)]. This compound was prepared as described above for
[CuII(H2L1)] by using Zn(ClO4)2‚6H2O (1.0 mmol) instead of Cu(ClO4)2‚
6H2O. Yield: 385 mg (67%). This yellow-brown compound is air
sensitive and was stored under an Ar atmosphere. 1H NMR (400 MHz,
CDCl3): δ 1.27 (s, 18H), 1.44 (s, 18H), 5.21 (broad s, 2H), 5.9 (very
broad, ∼1H), 6.69 (s, br, 2H), 6.90 (overlapping s, br, 4H), 7.14 (s, br,
2H). 13C{1H} NMR (100.6 MHz, CDCl3): δ 147.0, 142.6, 135.4, 128.8,
122.2, 120.4, 118.7, 117.5, 34.9, 34.4, 31.6, 29.7. Anal. Calcd for
C34H46N2O2Zn: C, 70.39; H, 7.99; N, 4.83; Zn, 11.27. Found: C, 70.11;
H, 7.98; N, 4.85; Zn, 10.99.
1
crystals of [CuII(L3)]‚CH3CN. Yield: 280 mg (49%). H NMR (400
MHz, CDCl3): δ 0.83-1.93 (m, 36H), 6.05-7.61 (m, 4H), 20.8 (s,
1H), 24.6 (s, 1H), 39.1 (s, 1H), -6.2 (s, 1H). Anal. Calcd for
C34H44N2O2Cu: C, 70.86; H, 7.70; N, 4.86. Found: C, 70.42; H, 7.75;
N, 4.83.
[CuII(L4)]PF6. To a solution of [CuII(L3)] (144 mg; 0.25 mmol) in
dry CH2Cl2 (30 mL) was added ferrocenium hexafluorophosphate (83
mg; 0.25 mmol) at -10 °C. After being stirred for 2 h, the violet clear
solution was stored at -20 °C for 12 h, after which time a violet
precipitate was collected by filtration. Yield: 40 mg (22%). Anal. Calcd
for C34H44N2O2CuPF6: C, 56.62; H, 6.15; N, 3.88; Cu, 8.81. Found:
C, 56.35; H, 6.14; N, 3.91; Cu, 8.74.
[CuII(L5)](ClO4)2. To a solution of [CuII(L3)] (288 mg; 0.5 mmol)
in dry CH2Cl2 (50 mL) at -20 °C was added [NiIII(tacn)2](ClO4)3 (615
mg; 1.0 mmol),9 whereupon the color of the solution changed to deep
red within 2 min and to carmine red after 6 h at -20 °C. After filtration,
this solution was kept at -80 °C for 10 h, during which time 30 mg of
a red-brown microcrystalline material precipitated, which was filtered
off at -50 °C and discarded. The resulting solution was stored at -80
°C for 48 h, during which time a red precipitate formed. Yield: 83 mg
Kinetic Measurements and Product Analyses. The kinetics of the
reactions of complexes [M(L4)]+ (M ) Cu, Zn) with methanol and
ethanol under anaerobic conditions in CH2Cl2 at 22 ( 1 °C were
measured spectrophotometrically using pseudo-first-order conditions
(excess alcohol). Pseudo-first-order rate constants were obtained from
plots of ln(At - A∞) vs time, where At is the absorbance at time t and
A∞ is the absorbance of the solution after completion of the reaction.
Such plots were linear for 4-5 half-lives.

9610 J. Am. Chem. Soc., Vol. 121, No. 41, 1999
Table 6. Crystallographic Data
Chaudhuri et al.
literature tables.24 All non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were placed at calculated positions and refined as
riding atoms with isotropic displacement parameters.
[Cu(L3)]‚CH3CN
[Zn(L3)]‚CH3CN
chem formula
fw
space group
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å
C36H47CuN3O2
617.31
P1h
C36H47N3O2Zn
619.14
P1h
Physical Measurements. Electronic spectra of complexes and
spectra of the spectroelectrochemical investigations were recorded on
a Perkin-Elmer Lambda 19 (range 220-1400 nm) and on a HP 8452A
diode array spectrophotometer (range 220-820 nm), respectively.
Cyclic voltammograms, square-wave voltammograms, and coulometric
experiments were performed on EG&G equipment (potentiostat/
galvanostat model 273A). Potentials were referenced vs the Ag/AgNO3
electrode, and ferrocene was added as internal standard. EPR spectra
of complexes were measured on a Bruker ESP 300E spectrometer
equipped with a helium flow cryostat (X-band, Oxford Instruments ESR
910). The solution spectra were simulated with the program written
by F. Neese,25 whereas the spin triplet spectra measured at low
temperatures were analyzed on the basis of a spin Hamiltonian
description of the electronic ground state:
10.156(1)
10.435(1)
16.845(2)
83.97(2)
75.47(2)
69.06(2)
1613.7(3)
2
10.173(1)
10.440(1)
16.856(2)
83.95(2)
75.41(2)
69.11(2)
1618.4(3)
2
Z
T, K
F
100(2)
1.270
7.12
15 017
7897/5784
392
58.0
0.0455
0.1019
100(2)
1.271
7.94
15 697
8533/6157
392
60.0
0.0483
0.1111
calcd, g cm-3
µ(Mo KR), cm-1
reflns collected
unique reflns/[I > 2σ(I)]
no. of parameters
2θmax, deg
He ) D[Sz2 - S(S + 1)/3 + (E/D)(Sx2 - Sy )] + µBB‚gj‚S (20)
2
R1a [I > 2σ(I)]
where S ) 1 is the spin of the coupled system and D and E/D are the
wR2b [I > 2σ(I)]
axial and rhombic zero-field parameters. These simulations were
5
performed with a program which was developed from the S )
/
2
a R1 ) ∑||Fo| - |Fc||/∑|Fo|. b wR2 ) [∑[w(Fo2 - Fc2)2]/∑[w(Fo2)2]1/2
,
routines of Gaffney and Silverstone26 and which specifically makes
use of the calculation of transition fields based on a Newton-Rapshon
iterative method as described there. Temperature-dependent magnetic
susceptibilities of powdered samples were measured by using a SQUID
magnetometer (Quantum Design) at 1.0 T in the range 2-300 K. For
calculations of the molar magnetic susceptibility, the raw data were
corrected for underlying diamagnetism of the sample by using tabulated
Pascal’s constants. NMR spectra were recorded on Bruker DRX 400
and DRX 500 instruments.
2
2
where w ) 1/σ2(Fo ) + (aP)2 + bP, P ) (Fo + 2Fc2)/3.
The kinetics of the catalytic reactions were measured under turnover
conditions by spectrophotometric determination of the formation of the
products (aldehyde and H2O2) using the initial rate method. In all cases,
the rates of formation of aldehyde and H2O2 were identical within
experimental error. The initial rates were obtained by quenching aliquots
of the reaction mixture after times t1, t2, ..., tn and analyzing their content
of products.
These aliquots were passed through a column containing the ion-
exchange resin Amberlyst 15 Dry (Supelco) in order to remove the
catalyst. H2O2 was determined spectrophotometrically as peroxotitanyl
species21 and, similarly, formaldehyde as 3,5-diacetyl-1,4-dihydrolu-
tidine10 and acetaldehyde with the reagent 3-methylbenzthiazoline-2-
on-hydrazone and FeCl3.10
Acknowledgment. We thank the Fonds der Chemischen
Industrie and the Degussa-Hu¨ls (Frankfurt, Germany) for
financial support. We thank F. Reikowski for recording the EPR
spectra.
Supporting Information Available: Figures S1-S3, show-
ing EPR spectra of [Cu(L3)], [Cu(H2L1)], and [Cu(L4)]PF6,
respectively, Figure S4, showing the temperature dependence
of the half-field EPR signal of [Cu(L4)]PF6 in frozen CH2Cl2
solution, and Figure S5, showing the EPR spectrum of [Cu-
(L5)](ClO4)2; tables of crystallographic and structure refinement
data, atom coordinates, bond lengths and angles, anisotropic
thermal parameters, and calculated positional parameters of H
atoms for complexes [M(L3)]‚CH3CN (M ) Cu, Zn) (PDF).
An X-ray crystallographic file, in CIF format, is also available.
This material is available free of charge via the Internet at
http://pubs.acs.org.
X-ray Crystallographic Data Collection and Refinement of the
Structures. Single crystals of [M(L3)]‚CH3CN (M ) Cu, Zn) were
selected from the mother liquor with a glass fiber and immediately
mounted on a Siemens SMART CCD detector diffractometer, equipped
with a cryogenic nitrogen cold stream to prevent loss of solvent.
Graphite-monochromated Mo KR radiation (λ ) 0.710 73 Å) was used.
Crystallographic data of the compounds are listed in Table 6. Cell
constants for 1 and 4 were obtained from a least-squares fit of the setting
angles of 5915 and 6182 strong reflections, respectively. Intensity data
were collected at -173(2) °C by a hemisphere run taking frames at
0.30° in ω. Data were corrected for Lorentz and polarization effects,
and a semiempirical absorption correction was carried out using the
program SADABS.22 The Siemens ShelXTL23 software package was
used for solution, refinement, and drawings of the structures. All
structures were solved and refined by direct methods and difference
Fourier techniques. Neutral atom scattering factors were obtained from
JA991481T
(24) International Tables for X-ray Crystallography; Kynoch Press:
Birmingham, U.K., 1991.
(25) Neese, F. Computer program EPR, DOS Version, University of
Konstanz, Germany, 1993.
(26) Gaffney, B. J.; Silverstone, H. J. Simulation of the EMR Spectra
of High-Spin Iron in Proteins. In Biological Magnetic Resonance; Berliner,
L. J., Reuben, J., Eds.; Plenum Press: New York, London, 1993; Vol. 13.
(21) Eisenberg, G. M. Anal. Chem. 1943, 15, 327.
(22) Sheldrick, G. M., Universita¨t Go¨ttingen, 1994.
(23) ShelXTL V5.03; Siemens Analytical X-ray Instruments, Inc.: Mad-
dison, WI, 1995.